The World Health Organization consolidated guidelines now recommend first-line treatment with dolutegravir (DTG)/tenofovir (TDF)/emtricitabine (FTC) or lamivudine (3 TC) for all HIV-infected adults.1 DTG has better tolerability, fewer treatment-limiting toxicities, higher barrier to resistance, and faster viral load (VL) suppression than efavirenz (EFV)-based antiretroviral treatment (ART).2,3 Maternal VL at delivery is a strong predictor of in utero and intrapartum mother-to-child transmission (MTCT) and DTG-based ART started in pregnancy has faster HIV RNA suppression than other ART regimens.4–12 These advantages have been cited as reasons for using DTG during pregnancy for the prevention of MTCT,13–17 although no direct comparisons of sufficient size have been performed to determine whether DTG offers an actual advantage for prevention of MTCT compared with EFV-based ART.
In 2016, Botswana was the first African country to change first-line ART from EFV/TDF/FTC to DTG/TDF/FTC, including for pregnant women. This change provided an opportunity for the first large-scale evaluation of MTCT risk for DTG-based ART (and the first evaluation in a programmatic setting). We used data from 2 studies that were ongoing in Botswana during the nationwide DTG rollout to compare in utero MTCT with either EFV- or DTG-based ART started in pregnancy.
The Early Infant Treatment Study
Between April 2015 and July 2018, the Early Infant Treatment Study (EIT) (NIH U01AI114235) screened HIV-exposed infants for in utero MTCT using qualitative DNA PCR at <96 hours of life. EIT screening occurred at approximately 20% of delivery facilities in Botswana, of which 4 hospitals (Nyangabgwe, Princess Marina, Scottish Livingstone, and Selebi Phikwe) accounted for 76% of screened infants. For all screened infants, the delivery site, date of birth, maternal age, infant gestational age, infant birth weight, infant sex, and maternal MTCT risk factors through maternal report were recorded. Maternal MTCT risk factors included: <8 weeks of ART in pregnancy (including no ART), poor ART adherence reported in pregnancy, maternal VL known to be >400 copies/mL at screening, maternal CD4 known to be <250 (or 250–350) at screening, and no maternal zidovudine (ZDV) receipt during labor. Infants were excluded from screening if they were hospitalized for severe medical illness or if they had a medical condition making it unlikely that they would survive.
Infants who screened positive for HIV on qualitative HIV-1 DNA PCR using Roche TaqMan v2 (Roche Diagnostics, Mannheim, Germany) were offered enrollment in EIT if mother/guardian was 18 years of age and able to provide additional informed consent for participation, and neonate gestational age at birth 35 weeks, birth weight 2000 grams, age <96 hours after birth, ability to initiate ART within 7 days after birth, and eligible for ART through the Botswana government program. For infants enrolled into the EIT study, detailed maternal ART history and CD4 count and VL data were obtained.
The Tsepamo Study
For infants who screened HIV-negative and were not enrolled in the EIT Study, maternal ART regimen and start date were not available within the EIT study, but were available for those who could be linked to the Tsepamo Birth Outcomes Surveillance Study18 (NIH R01HD080471). The Tsepamo Study provided separate surveillance of maternal obstetric records for about 45% of all births in Botswana during the EIT screening period, including at the 4 EIT sites where most screenings occurred. From November 2015 to July 2018, most screened EIT infants became linkable with the Tsepamo database if delivery occurred at a Tsepamo surveillance site.
EIT MTCT risk factor analysis was performed using Fisher exact test, proportion confidence intervals (CIs) were estimated using Clopper–Pearson for the binomial proportions, risk difference using Newcombe CI, and exact logistic regression to calculate odds ratios (OR). Comparisons between ART regimens were performed using Fisher exact test and Wilcoxon rank sum test; exact logistic regression was used to calculate the OR for differences in MTCT.
All mothers of screened children for EIT signed written consent approved by ethical review boards in Botswana (Health Research Development Committee) and Boston (Harvard T.H. Chan School of Public Health Office of Human Research Administration), and both studies were approved by these same review boards.
Infants Screened for HIV in EIT
The EIT study screened 10,622 infants and identified 42 (0.40%) who tested HIV+ at <96 hours of age. Of the 10,622 screened, 10,387 (97.8%) were exposed to some type of ART in utero; of these, MTCT occurred among 24 (0.23%). The majority (26 infants, 62%) of the 42 HIV+ infants were screened at Princess Marina Hospital (13), Nyangabgwe Hospital (9), and Selebi Phikwe Hospital (4). The remaining were screened at the following hospitals/clinics: Area W Clinic (2), Athlone Hospital (1), BH3 Clinic (1), Botshabelo Clinic (1), GWest Clinic (2), Kanye Hospital (1), Masunge Hospital (1), Oodi Council Clinic (1), Scottish Livingston Hospital (1), Tatisiding Clinic (1), Tonota Clinic (1), and Tutume Hospital (1); and 2 infants were screened at their houses.
MTCT Risk Factors in Infants Screened for HIV in EIT
Among all infants screened at birth for HIV by the EIT study, 8532 (80.3%) had no MTCT risk factors. However, among the 42 HIV+ infants, 35 (83.3%) had at least one identifiable MTCT risk factor at the time of screening (before laboratory information available through enrollment in the study). In multivariate analysis of all risk factors assessed at the time of infant screening, <8 weeks of ART in pregnancy (including no ART), poor ART adherence during pregnancy, and detectable maternal VL at last test (when available) were all significant independent predictors of HIV transmission (Table 1). Of note, there were 5 transmissions with no associated risk factors for transmission at the time of screening, all of which occurred among women on DTG. However, after further laboratory testing at EIT enrollment, 3 of these women were identified as having both HIV RNA >40 copies/mL and CD4 <250 cells/mm3, thus leaving only 2 mothers with no identifiable risk factors upon detailed review.
Maternal and Infant HIV-1 RNA at Delivery in the EIT Study
Among the 42 HIV+ infants at screening, 40 were enrolled in the EIT study (2 infants did not consent for enrollment). Seventeen (42.5%) of the 40 infants had no in utero ART exposure, 2 (5.0%) were exposed to LPV/r-based regimens, 10 (25.0%) to EFV/TDF/FTC, and 11 (27.5%) to DTG/TDF/3 TC. The median maternal age was 27.0 years (interquartile range 22.0, 29.5 years), 23 (57.5%) mothers were primigravids, 38 (95.0%) mothers were single, and 27 (67.5%) mothers were unemployed. Among the 23 HIV+ infants exposed to any ART in utero, the mothers of 19 (83%) had started the ART in pregnancy, at a median of 29 weeks gestational age; the mothers of the other 4 infants started before conception (Table 2).
Among the mothers of enrolled infants, maternal VL at delivery was available for all mothers. Median maternal VL for mothers receiving DTG (n = 11) was significantly lower than those on EFV, other ART, or no ART (n = 29); median maternal VL at delivery was 56 copies/mL (range <40, 85,697) for the DTG group, compared with a median maternal VL at delivery of 10,259 copies/mL, 29,085 copies/mL, and 64,072 copies/mL for EFV, other ART, and no ART, respectively (P < 0.01). Of the 40 mothers, 35 (87.5%) mothers had HIV RNA >40 copies/mL at delivery. The 5 transmitting women with HIV RNA <40 copies/mL at delivery were all receiving DTG, started from 22 to 35 weeks gestational age (at 22, 23, 30, 30, and 35 weeks). Transmitting women on DTG were significantly more likely to have HIV RNA <40 copies/mL at delivery than transmitting women on other ART regimens; 5 (45%) of 11 transmitting mothers taking DTG had HIV RNA <40 copies/mL at delivery, versus 0 (0%) among 12 transmitting mothers taking EFV or other ART regimens (P = 0.01).
Among the 39 infected infants with documented baseline HIV RNA, infants born to women on DTG (n = 11) had lower HIV RNA at enrollment to EIT than those exposed to no in utero ART (n = 17); median HIV RNA was 310 copies/mL (range 79, 389,270 copies/mL) for the DTG group, compared with a median HIV RNA 31,708 copies/mL (range <40, >10,000,000 copies/mL) for those with no in utero ART exposure (P = 0.04). In contrast with infants on in utero DTG vs. no ART, no differences in baseline HIV RNA were found between infants born to women receiving either EFV or any other ART regimen as compared with no ART (P = 0.18 and P = 0.95, respectively) (Table 2). Similarly, the proportion of infants with HIV RNA <400 copies/mL at enrollment was greater for those exposed to DTG (63.6%) than for all other infants, including any other ART (0.0%) (P = 0.004) or no ART (17.6%) (P = 0.02).
Infants Screened for EIT and Linked to Tsepamo Surveillance Data
In total, 5064 (47.7%) infants screened for EIT could be linked to maternal ART information in the Tsepamo Surveillance database. Reasons for inability to link EIT screens to Tsepamo included delivering/screening at a non-Tsepamo site (2,538, 45.7%) and lack of capture of the screening identifier in Tsepamo at the time of data extraction (3,020, 54.3%). Linked and unlinked infants had similar maternal MTCT risk factors and occurred without marked difference over calendar time of study participation (data not shown). Among linked infants, 1235 (24.4%) were exposed to DTG-based ART, 2411 (47.6%) to EFV-based ART, 1246 (24.6%) to other ART, 37 (0.73%) to multiple ART regimens, and 135 (2.7%) to no ART (Fig. 1). There was a nonsignificantly (P = 0.12) higher proportion of linkages among HIV+ infants (60%) compared with all screened infants (48%). For the 5064 linked infants, the 25 transmissions were distributed by hospital site as follows: Princess Marina Hospital (12%, 48%), Nyangabgwe (8%, 32%), Selebi Phikwe (4%, 16%), and Scottish Livingstone Hospital (1%, 4%). Complete distribution of linked infants and in utero antiretroviral regimen by site are shown in Table 1, Supplemental Digital Content, https://links.lww.com/QAI/B450.
Direct comparisons were performed for DTG and EFV exposures. Compared with mothers on EFV, mothers on DTG were slightly younger (mean 28.6 vs. 30.3 years, P < 0.001) and more commonly primigravid (24.9% vs. 14.0%, P < 0.0001) (Table 3). Compared with EFV, mothers on DTG were mostly diagnosed with HIV during pregnancy (61.4% vs. 25.1%, P < 0.0001) and were more likely to have had their ART regimen started during pregnancy (82.4% vs. 37.1%, P < 0.0001). For mothers starting ART during pregnancy, DTG was initiated slightly earlier in pregnancy than EFV (median 19 weeks vs. 20 weeks).
MTCT With DTG and EFV Started Preconception and in Pregnancy
In total, 17 linked HIV+ infants were exposed to either DTG (n = 8) or EFV (n = 9) in utero. Of 213 infants whose mothers initiated DTG before pregnancy, none were infected (0.00%, 95% CI: 0.00% to 1.72%). Of the 1497 infants whose mothers initiated EFV before pregnancy, 1 (0.07%, 95% CI: 0.00% to 0.37%) was infected. Among infants whose mothers started either DTG or EFV in pregnancy, transmissions occurred in 8/999 (0.80%, 95% CI: 0.35% to 1.57%) exposed to DTG and 8/883 (0.91%, 95% CI: 0.39% to 1.78%) exposed to EFV (OR 0.88, 95% CI: 0.29 to 2.71) (Table 3). The absolute MTCT risk difference when starting either regimen in pregnancy was 0.11% (95% CI: −0.79% to 1.06%). Most MTCT events (4 of 8 with DTG, 6 of 9 with EFV) occurred among women starting ART <90 days before delivery (Table 4).
We performed the first assessment of in utero MTCT in a national program using DTG-based ART, and found no difference in the low overall MTCT risk between DTG-based ART and EFV-based ART. These findings are consistent with the high ART coverage during pregnancy achieved in Botswana.19,20 We also identified low HIV RNA in mothers and infants, and fewer identifiable risk factors for MTCT, with DTG use during pregnancy.
The World Health Organization consolidated guidelines recommend first-line treatment with DTG for HIV-infected adults, including pregnant women.21 DTG has a higher barrier to resistance and faster HIV RNA suppression than EFV,2 and some experts have predicted that these favorable factors may also translate into lower MTCT.13,14,17 However, although our study demonstrated low maternal HIV RNA at delivery with DTG, there was no evidence for improvement in the already very low in utero MTCT rate seen with EFV among women started on ART in pregnancy, and we were able to exclude a difference greater than 1.06% with 95% confidence. Although there are no large comparator studies yet for DTG, lower in utero MTCT rate has been documented with EFV-based ART in Botswana when compared with other settings,19,22–24 and this low rate was also seen in our study. Although this study presents the largest series of DTG exposures in pregnancy to date, CIs were wide and we cannot exclude higher transmission than the point estimates. It remains unknown whether DTG may reduce MTCT in settings with higher overall transmission. Two randomized controlled trials compare HIV RNA at delivery as a surrogate for MTCT risk for women initiating DTG vs EFV.25 Initial data from one trial (DolPHIN-2) demonstrated increased speed of viral suppression in the DTG arm, and low MTCT in both arms, which is consistent with our findings.26 Our study adds to these clinical trial data by providing greater power to directly compare MTCT for each regimen in a national program setting.
Most transmitting woman in our study had at least one identifiable risk factor for MTCT during pregnancy. Consistent with prior studies,22,27–29 a previous analysis of maternal risk factors for infants screened in the EIT study between April 2015 to April 2016 (conducted before DTG usage) found that all had at least one risk factor (either <8 weeks of maternal ART or known lack of maternal VL suppression).30 However, with extension of EIT to July 2018, we found that 2 (18%) of 11 HIV-infected infants exposed to in utero DTG were infected without identifiable risk factors, despite the fact that the prevalence of risk factors did not change substantially overall among screened infants. The significance of this finding is unclear, but warrants further study. We cannot exclude unrecognized viremia for these transmissions, but cases such as this add to our knowledge about whether “undetectable equals untransmittable” (U=U) for MTCT, and at present (and until we have trials with more precise VL monitoring), such a claim unfortunately cannot be made.31
The observation that maternal DTG use in pregnancy led to low maternal and infant HIV RNA at delivery, if supported by future data, has several implications. First, it suggests that HIV RNA at the time of delivery may be a less reliable marker for in utero MTCT events when women start DTG in pregnancy compared with historical ART regimens.13,26 Second, it has diagnostic implications, with a possibility for a higher proportion of infants exposed to DTG in utero having lower than quantifiable levels of baseline HIV RNA (and possibly DNA) that could result in false-negative screening results (especially if HIV RNA screening is performed with a cutoff of <400, as has been considered in some settings).32,33 Finally, if infant HIV RNA is directly reduced by placental transfer of maternal DTG, there may be beneficial effects for HIV+ infants, who essentially begin treatment in utero and may have reduced viral reservoir.
Strengths of this study include the large population of HIV-exposed infants within a demographically homogeneous study population outside of a clinical trial. The linkage with the Tsepamo surveillance allowed for ART exposure data for a very large number of DTG-exposed infants (including gestational week of ART initiation). In addition, the study had limited selection bias of infants who participated in the study because nearly all infants who met inclusion criteria were screened (very few mothers declined to participate) and >95% of deliveries in Botswana occur in hospital. Limitations of the study include the linkage of a screening study with an observational surveillance study, thus leading to possible confounding factors for MTCT by regimen if unmeasured differences between linked and unlinked infants were present. This may have led to a small, nonsignificant overestimate of MTCT in the linked population (0.49%) compared with the overall screened EIT population (0.40%). Although maternal demographic data and other factors were similar, because DTG was introduced as first-line ART after EFV, there was possible temporal confounding between the 2 treatment periods. Because there were limited HIV RNA data during pregnancy, including before ART initiation, time to viral suppression could not be measured. In addition, in the EIT screening data, there could be unrecognized confounding by risk factors that were not identified, and because of the low number of VL and CD4 tests available for analysis. Finally, our study only measured in utero MTCT detectable in the first week of life. Although intrapartum MTCT has become rare in the ART era,34 we can make no conclusion as to whether the more rapid HIV RNA decline with DTG may be of benefit for intrapartum transmission.
In conclusion, this study provides the largest analysis to date of MTCT risk with DTG use in pregnancy. Transmitting mother/infant pairs exposed to DTG had significantly lower VLs at delivery/birth than those exposed to EFV or other ART regimens. Two women on DTG who transmitted did so without apparent MTCT risk factors, and these findings warrant further evaluation. Reassuringly, MTCT has remained very low among women receiving DTG in Botswana. Although we were not powered to compare MTCT by ART regimen, no significant MTCT differences were observed between DTG-based and EFV-based ART, and the MTCT risk seemed similar. As in prior studies, MTCT risk was highest for both groups when ART was started in the third trimester, thus highlighting the importance of early HIV diagnosis, preferably before pregnancy, and early ART initiation.
1. Transition to New Antiretroviral Drugs in HIV Programmers: Clinical and Programmatic Considerations. Geneva, Switzerland: World Health Organization; 2017.
2. Hill A, Clayden P, Thorne C, et al. Safety and pharmacokinetics of dolutegravir in HIV-positive pregnant women: a systematic review. J Virus Erad. 2018;4:66–71.
3. Kanters S, Vitoria M, Doherty M, et al. Comparative efficacy and safety of first-line antiretroviral therapy for the treatment of HIV infection: a systematic review and network meta-analysis. Lancet HIV. 2016;3:e510–e520.
4. O'Shea S, Newell ML, Dunn DT, et al. Maternal viral load, CD4 cell count and vertical transmission of HIV-1. J Med Virol. 1998;54:113–117.
5. Leroy V, Montcho C, Manigart O, et al. Maternal plasma viral load, zidovudine and mother-to-child transmission of HIV-1 in Africa: DITRAME ANRS 049a trial. AIDS Lond Engl. 2001;15:517–522.
6. Warszawski J, Tubiana R, Le Chenadec J, et al. Mother-to-child HIV transmission despite antiretroviral therapy in the ANRS French Perinatal Cohort. AIDS Lond Engl. 2008;22:289–299.
7. Stewart RD, Wells CE, Roberts SW, et al. Benefit of interpregnancy HIV viral load suppression on subsequent maternal and infant outcomes. Am J Obstet Gynecol. 2014;211:297e1–6.
8. Mock PA, Shaffer N, Bhadrakom C, et al. Maternal viral load and timing of mother-to-child HIV transmission, Bangkok, Thailand. Bangkok collaborative perinatal HIV transmission study group. AIDS Lond Engl. 1999;13:407–414.
9. Sripan P, Le Coeur S, Amzal B, et al. Modeling of in-utero and intra-partum transmissions to evaluate the efficacy of interventions for the prevention of perinatal HIV. PLoS One. 2015;10:e0126647.
10. Ewing AC, Ellington SR, Wiener JB, et al. Predictors of perinatal HIV transmission among women without prior antiretroviral therapy in a resource-limited setting: the breastfeeding, antiretrovirals and nutrition study. Pediatr Infect Dis J. 2019;38:508–512.
11. Landes M, van Lettow M, Nkhoma E, et al. Low detectable postpartum viral load is associated with HIV transmission in Malawi's prevention of mother-to-child transmission programme. J Int AIDS Soc. 2019;22:e25290.
12. Plessis NMD, Muller CJB, Avenant T, et al. An early infant HIV risk score for targeted HIV testing at birth. Pediatrics. 2019;143:e20183834.
13. Schalkwijk S, Greupink R, Colbers AP, et al. Placental transfer of the HIV integrase inhibitor dolutegravir in an ex vivo human cotyledon perfusion model. J Antimicrob Chemother. 2016;71:480–483.
14. Dugdale CM, Ciaranello AL, Bekker LG, et al. Risks and benefits of dolutegravir- and efavirenz-based strategies for South African women with HIV of child-bearing potential: a modeling study. Ann Intern Med. 2019;170:614–625.
15. Bornhede R, Soeria-Atmadja S, Westling K, et al. Dolutegravir in pregnancy-effects on HIV-positive women and their infants. Eur J Clin Microbiol Infect Dis. 2018;37:495–500.
16. Pinnetti C, Tintoni M, Ammassari A, et al. Successful prevention of HIV mother-to-child transmission with dolutegravir-based combination antiretroviral therapy in a vertically infected pregnant woman with multiclass highly drug-resistant HIV-1. AIDS Lond Engl. 2015;29:2534–2537.
17. Phillips AN, Venter F, Havlir D, et al. Risks and benefits of dolutegravir-based antiretroviral drug regimens in sub-Saharan Africa: a modelling study. Lancet HIV. 2019;6:e116–e127.
18. Zash R, Holmes L, Diseko M, et al. Neural-tube defects and antiretroviral treatment regimens in Botswana. N Engl J Med. 2019;381:827–840.
19. Progress Report of the National Response to the 2011 Declaration of Commitments on HIV and AIDS. Gaborone, Botswana: National AIDS Coordinating Agency, Republic of Botswana; 2015.
20. Ending AIDS: Progress towards the 90-90-90 Targets. Geneva, Switzerland: Joint United Nations Programme on HIV/AIDS; 2017.
21. WHO | Transition to new antiretroviral drugs in HIV programmes: clinical and programmatic considerations. WHO. Available at: http://www.who.int/hiv/pub/toolkits/transition-to-new-arv-technical-update/en/
. Accessed August 8, 2019.
22. Myer L, Phillips TK, McIntyre JA, et al. HIV viraemia and mother-to-child transmission risk after antiretroviral therapy initiation in pregnancy in Cape Town, South Africa. HIV Med. 2017;18:80–88.
23. Dooley KE, Denti P, Martinson N, et al. Pharmacokinetics of efavirenz and treatment of HIV-1 among pregnant women with and without tuberculosis coinfection. J Infect Dis. 2015;211:197–205.
24. Hoffman R, Black V, Technau K, et al. Effects of highly active antiretroviral therapy duration and regimen on risk for mother-to-child transmission of HIV in johannesburg, South Africa. J Acquir Immune Defic Syndr. 2010;54:35–41.
25. Bailey H, Zash R, Rasi V, et al. HIV treatment in pregnancy. Lancet HIV. 2018;5:e457–e467.
26. Kintu K, Malaba T, Nakibuka J, et al. RCT of dolutegravir vs efavirenz-based therapy initiated in late pregnancy: dolphin-2. In. Seattle, WA: CROI Abstract Publications; 2019.
27. Wang Q, Wang L, Fang L, et al. Timely antiretroviral prophylaxis during pregnancy effectively reduces HIV mother-to-child transmission in eight counties in China: a prospective study during 2004–2011. Sci Rep. 2016;6:34526.
28. Momplaisir FM, Brady KA, Fekete T, et al. Time of HIV diagnosis and engagement in prenatal care impact virologic outcomes of pregnant women with HIV. PLoS One. 2015;10:e0132262.
29. Lima YA, Cardoso LP, Reis MN, et al. Incident and long-term HIV-1 infection among pregnant women in Brazil: transmitted drug resistance and mother-to-child transmission. J Med Virol. 2016;88:1936–1943.
30. Ibrahim M, Maswabi K, Ajibola G, et al. Targeted HIV testing at birth supported by low and predictable mother-to-child transmission risk in Botswana. J Int AIDS Soc. 2018;21:e25111.
31. Shapiro RL, Rossi S, Ogwu A, et al. Therapeutic levels of lopinavir in late pregnancy and abacavir passage into breast milk in the Mma Bana Study, Botswana. Antivir Ther. 2013;18:585–590.
32. Lee BE, Plitt SS, Jayaraman GC, et al. Use of quantitative HIV RNA detection for early diagnosis of HIV infection in infants and acute HIV infections in alberta, Canada. J Clin Microbiol. 2012;50:502–505.
33. Mazanderani AH, Moyo F, Kufa T, et al. Brief report: declining baseline viremia and escalating discordant HIV-1 confirmatory results within South Africa's early infant diagnosis program, 2010-2016. J Acquir Immune Defic Syndr. 2018;77:212–216.
34. Rollins N, Mahy M, Becquet R, et al. Estimates of peripartum and postnatal mother-to-child transmission probabilities of HIV for use in Spectrum and other population-based models. Sex Transm Infect. 2012;88(suppl 2):i44–51.